Running jobs on CARC Systems - Workshop

This repository contains preconfigured slurm job script examples, to run on USC CARC Discovery Cluster.

By participating in this workshop, you’ll:

• Learn the fundamental skills and tools required to run jobs on CARC Cluster

• Gain experience with different job types and submission file structure

• Take a brief look into troubleshooting your jobs

• Build confidence to create your own jobs

Duration: 1.5 hours

Prerequisites: An understanding of fundamental CARC Cluster setup and CARC Ondemand.

Suggested materials to satisfy prerequisites:

• Gettting started with Discovery

• CARC Ondemand

Hardware Requirements: Desktop or laptop computer capable of running the latest version of Chrome or Firefox.

Language: English

Getting Started

To get started, use a web browser and log on to USC OnDemand Instance at https://ondemand.carc.usc.edu . You need to be on USC Network and need to your USC credentials to log in. More details on how to logon to USC Open OnDemand can be found at our guide

To start a Jupyter notebook server, Click on Interactive Apps and then select JupyterLab.

When launching the Jupyter Lab, it is important to select the following

• For Cluster, specify Discovery
• For Account, specify the account you normally use, like ttrojan_123
• For Partition, specify main
• Set the number of CPUs to 1
• Set Memory to 2GB
• Set the number of hours to 4

For the CARC Summer bootcamp 2024, please do the above with the following change:

• For Account, specify osinski_982

For the "DSR626" course, please do the above with the following change:

• For Account, specify ujadhav_1068

The above settings are important to ensure you start a JupyterLab Server on a node that is located on the Discovery cluster and has slurm management engine installed and configured

Once your JupyterLab Session starts, Click Connect to JupyterLab button

From there, Click on File -> New and then click on Terminal to get the terminal

Once in the terminal, clone this GitHub Repository

module load usc git
git clone https://github.com/uschpc/running-jobs-on-CARC-systems.git

In Jupyter, navigate to the example you are interested in, and step through the notebook.

For first-time users, we highly recommend to do the notebooks in order, as they build up on concepts in the previous notebooks.